my rootvg is varied off!!!
 

I got a great problem!
After reboot a node of my cluster seems to be up (rootvg fs are mounted, tcpip services are active) but rootvg results not varied on. My rootvg is defined on hdisk0 and hdisk1 (mirrored) and other vg occupies from hdisk2 up to hdisk9. Now the system sees the inexistent hdisk10 and hdisk11 defined but not assigned to a vg.

hdisk0 and hdisk1 are been removed manually, before this operation the lspv command output showed 12 disk with hdisk0 and hdisk11 not assigned to a vg.

After rmdev -dl hdisk0 & rmdev -d hdisk1 lspv show this ouput:

Any ideas?
TNX

How did this happen? What you need to do is rmdev -dl hdisk10 and 11, then make sure HA is stopped on this node (smitty cl_stop) and run cfgmgr. hdisk0 & 1 should reappear. If they still say none for the volume group, do importvg -y rootvg hdisk0, then try lspv again. How are your rootvg filesystems mounted if there is no rootvg?

This happened after a shutdown. Maybe there was a problem with the sysplanar or with the scsi bus, anyway i can't remove hdisk10 and hdisk11 because the system sees them as busy. I think it's an ODM problem: hdisk10 and hdisk11 are the same as hdisk0 and hdisk1 but not for the system. Now the situation is that the system booted from hdisk10 but the system does not recognize it as belonging to rootvg (PVID are different), so rootvg can't be varied on even if everything seems ok. Very interesting problem, but maybe it's faster to reinstall the system from a backup.

Definitively the problem is somewhere else. If rootvg is not varied on, /usr is not mounted, so you have no Unix system. Your rootvg is necesserily varied on. Of course this does not solve your problem of willing to know how to get the things fixed, nevertheless you can be sure that rootvg is varied on.

I agree with you zorba, rootvg must be varied on, and i think it is, anyway when i edit any command on rootvg (ie lsvg -l rootvg) i get the message:

"0516-010 : Volume group must be varied on; use varyonvg command".

Editing lspv i have no PV assigned to rootvg (rootvg is not listed at all) and hdisk10 and hdisk11 are not assigned to vg. Moreover PVID of hdisk10 and hdisk11 are both different from the PVID of hdisk0 and hdisk1, and i don't know how this can be possible if the disks are the same!
I think that somewhere in ODM (i don't know where!) there is an istance that makes the system identifies hdisk10 and hdisk11 as belonging to rootvg and somewhere else they are identified as new disks. File system are mounted and i tried to read from /usr and nmon says that i'm reading from hdisk10!

Something must be out of sync somewhere.
There is something absolutely not risky to be done : just use "smit vg", "activate the volume goup", if there is an error message do what is said (probably syncvg odm) and that's probably all what has to be done.
Nevertheless, having no disk seen as rootvg from the lspv output remains rather strange.